|Publication number||US7710942 B2|
|Application number||US 10/584,097|
|Publication date||May 4, 2010|
|Filing date||Dec 14, 2004|
|Priority date||Dec 23, 2003|
|Also published as||CN1898881A, CN100505572C, EP1700388A1, US20070160119, WO2005064810A1|
|Publication number||10584097, 584097, PCT/2004/4218, PCT/IB/2004/004218, PCT/IB/2004/04218, PCT/IB/4/004218, PCT/IB/4/04218, PCT/IB2004/004218, PCT/IB2004/04218, PCT/IB2004004218, PCT/IB200404218, PCT/IB4/004218, PCT/IB4/04218, PCT/IB4004218, PCT/IB404218, US 7710942 B2, US 7710942B2, US-B2-7710942, US7710942 B2, US7710942B2|
|Inventors||Emmanuel Ardichvili, Martial Gander|
|Original Assignee||St-Ericsson Sa|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (6), Classifications (10), Legal Events (3)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The present invention relates to a method of performing time drift compensation in a receiver and a receiver for performing time drift compensation.
In UMTS (Universal Mobile Telecommunications Service) data processing, data pertaining to different propagation paths reach a mobile station at different instants of time. Moreover, said data may reach the mobile station from the same or different transmitting sources, i.e. base stations. The information transmitted in each path in the air interface is sampled, demodulated and delayed at the receiver, and then combined with the correspondingly processed information of the other paths to improve the quality of the received signals.
In CDMA systems, spreading is used to translate each symbol into a sequence of bits. At the transmitter, each symbol is multiplied with a wideband spreading code. The higher the bandwidth used to transmit a signal, the lower the required signal-to-noise rate must be. When a higher bandwidth is employed, the signal is spread over a greater spectrum, and that makes the signal more tolerant to interference. The spreading code is also known as the channelization code and is used to distinguish the different channels. The spreading code comprises N bits, or chips, per symbol. Each symbol to be transmitted is converted into a longer sequence of chips, by means of the multiplication. The length of the spreading code, i.e. the number of chips comprised in each spreading code, is known as the spreading factor and is dependent on the channel type. At the receiver, the original symbols can be reconstructed by multiplying the received sequence of chips with a corresponding despreading chip sequence. When the chip sequence has been demodulated at the receiver, and the corresponding symbols thus have been reconstructed, the symbols of each path are delayed with respect to the last path received, and this will compensate for the delay of each path. Note that the receiver comprises one “finger” (branch) for each path to be processed. The paths are thus time-aligned to each other and combined to form a composite symbol that is expected to have a higher signal-to-noise ratio (SNR) than a symbol received in a single path. In order to combine the different paths meaningfully, channel parameters such as number of paths, delays and attenuation of each path must be known. A major problem in particular is that the characteristics of the paths vary with time, due to the fact that the receiver is mobile. These varying characteristics must be adjusted dynamically, and that is a highly complex process.
When combining the symbols by summing each path, another problem arises. Assume that a symbol is received from a first base station (BS1) and has to be delayed and time-aligned with a subsequently received symbol from a second base station (BS2). The delay T is represented by an integral number (I) of symbol periods and a variable delay portion (D) for adjusting the alignment of symbols. If the mobile moves towards BS1 while remaining at the same distance from BS2, the delay applied to the symbol originating from BS1 should be increased to keep the two symbols aligned due to this time drift. To align a symbol received from BS1 at time t to a symbol received from BS2 at time t+D, it is necessary to adjust (delay) the symbol of BS1 with the amount D. The symbols of BS1 may then be aligned with the symbols of BS2 such that the boundary of a symbol relating to BS1 is aligned with the boundary of a symbol relating to BS2. However, a problem may still arise: after the symbol of BS1 has been subject to the delay D, a BS1 symbol of order K+1 may be aligned with a BS2 symbol of order K, and not with a BS2 symbol of order K+1, which is required to produce a coherently composite symbol. The symbol of BS1 must then be delayed by I, i.e. one symbol period, in order for the BS1 symbol of order K to become aligned to the BS2 symbol of order K. The total delay T is thus I+D in this specific example. This is known as symbol alignment. Thereafter, symbols of each path may be coherently combined to create a composite symbol.
As the mobile moves towards BS1, the delay will increase and, consequently, the delay T must be increased by 2ÎI, 3ÎI, 4ÎI etc. Note that the adjustment delay D varies between zero and one symbol period I and must continuously be applied to the path of BS1. When implementing the delay adjustment in the mobile phone, delay line units are employed, one for each path. To comply with real-time requirements of the system in which the mobile is comprised, there is an upper limit to the amount of delay that each delay line unit can be allowed to insert.
This maximum delay is referred to as delay line maximum length, DLM. When the delay of two multi-path symbols exceeds this maximum delay, symbol alignment is no longer possible, and symbols relating to the path of BS1 will consequently be lost. This has the effect that there can be no combination of symbols from BS1 with symbols from BS2. Thus, the worst case scenario in this example is that only symbols relating to BS2 will be received.
The delay line unit of the path relating to BS1 will then be “reset”, which means that if the DLM is e.g. 5ÎI, the new value of the delay is set to 3ÎI and again, another attempt to align the path relating to BS1 with the path relating to BS2 will be undertaken, this time with new timing (delay) parameters. The resetting of the delay line has the effect that the symbols in delay line positions 4ÎI and 5ÎI are lost.
A more preferable way to perform the symbol alignment is to, once again, to make an adjustment of the path pertaining to BS1 by applying the delay D such that the boundary of a symbol relating to BS1 is aligned with the boundary of a symbol relating to BS2. Thereafter, as in the previous example, a BS1 symbol of order K+1 may be aligned with a BS2 symbol of order K, and not with a BS2 symbol of order K+1, which is required to produce a coherently composite symbol. The symbol of BS1 must then be delayed by I, i.e. one symbol period, in order for the BS1 symbol of order K to become aligned with the BS2 symbol of order K. However, if the maximum delay is reached, a coherent adding of symbols can no longer be accomplished. Therefore, both paths are subject to delays, and when the maximum delay is reached, the symbol of order K is omitted in both paths. This has the practical effect that the path of BS1 can be delayed one more symbol period at the expense of a completely lost symbol of order K (and of course at the expense of a slight delay of received symbols for the path pertaining to BS2 as these are no longer received in real-time). The omitting of symbols is referred to as time drift compensation.
If the mobile moves away from BS1 while remaining at the same distance from BS2, the delay applied to the symbols originating from BS1 should be decreased to keep the paths aligned due to this time drift. This case is symmetrical to the case where a symbol is omitted in both paths. Consequently, the symbol of order K relating to BS1 will not be omitted, but employed twice, i.e. summed twice with the symbol of order K pertaining to BS2.
In practice, not only two, but N different paths are coherently summed. Two major problems arise when combination of symbols is to be undertaken. First, the instant in time when symbols are to be omitted depends on path timing characteristics, and will vary for all paths. Second, in the above examples, it was assumed that BS2 was the reference while the timing of BS1 was adjusted. However, the symbols derived from BS2 may disappear due to fading, which has the effect that timing control and decisions taken to omit/repeat symbols must be done in relation to a new reference path. These problems will make the management of time-drift compensation difficult.
Implementation of a common delay line located before the demodulators and processing units has been described in French patent application No. 02 10452 of 21 Aug. 2002. The main difference compared to the previous delay line implementations is that, in this case, the delays are applied to the chips rather than to the symbols. This has a major impact on performance loss of the receiver, since a chip is omitted instead of a complete symbol, i.e. only 1/N out of a symbol consisting of N chips is omitted/repeated by this operation. The delay adjustments will occur at a factor N more occasions than in the case where the delays are applied on a symbol level, but these N compensations are performed on different symbols, so the modifying of chips is spread over N symbols, which minimizes the performance loss. Another major difference is that, at the output of the delay line, the fingers of the receiver are positioned to compensate for the propagation delays, so that the chips that originate from the respective propagation paths are output at the same instant of time. This has the effect that the position of each finger is changed simultaneously when a chip is omitted/repeated. This simplifies the software management as compared to prior art not employing a common delay line.
However, the receiver disclosed in French patent application No. 02 10452 of 21 Aug. 2002 still presents unacceptable performance losses. Another problem is that the software management, even though it has been simplified, is still rather complex.
An object of the present invention is to mitigate the above described problems and provide an efficient way of performing time drift compensation in a receiver and a set of dedicated resources to be implemented in the receiver to effect the time drift compensation.
A further object of the present invention is to provide improved performance gain of the receiver and simplified software management.
These objects are attained by a method of performing time drift compensation in a receiver according to claim 1 and a receiver for performing time drift compensation according to claim 10.
According to a first aspect of the invention, a method is provided in which a signal is received. Then, a control pulse is produced after a certain number of chips of the received signal has been received. A variable delay is applied to the received signal. The resulting delayed signal is provided to demodulation units in the receiver, in which delayed signal chips have been omitted or duplicated on the basis of the control pulse. A compensation signal is supplied to the demodulation units in the receiver, and this compensation signal indicates whether chips have been omitted or duplicated in the delayed signal. Finally, the delayed signal is demodulated such that the demodulation units consider the omission or duplication of chips in the delayed signal.
According to a second aspect of the invention, a receiver comprising a timer, a common delay unit and demodulation units is provided. The timer is arranged to produce a control pulse after having received a certain number of chips of a received signal. The common delay unit is arranged to apply a variable delay to the received signal and thus provide a delayed signal to the demodulation units in the receiver, in which delayed signal chips have been omitted or duplicated on the basis of the control pulse. Further, the timer is arranged to supply to the demodulation units in the receiver a compensation signal that indicates whether chips have been omitted or duplicated in the delayed signal. Moreover, the demodulation units are arranged to demodulate the delayed signal such that the demodulation units consider the omission or duplication of chips in the delayed signal.
The basic idea of the invention is that a receiver receives signals that originate from a transmitter. Before transmission, the transmitter translates each radio symbol into a sequence of bits by means of a spreading code. These bits are typically referred to as chips. These chips are then sent across an air interface and are subject to multipath propagation. In practice, this means that a given signal will, via a number of paths, arrive at the receiver at different instants of time due to different physical lengths of the paths. In prior art, this has been a problem and has caused undesired effects such as fading, delays and interference at the receiver. However, as previously described, it is also possible to benefit from multipath propagation by applying time drift compensation to the received signals, in order to eliminate the effects of internal delays among the received signals.
When a signal is received, i.e. a signal pertaining to one transmission path, a variable delay is applied to the signal in order to enable alignment of the delayed signal to a subsequently received signal. After having received a certain number of chips, a control pulse is produced, and this control pulse controls the variable delay applied to the received signal. The delay is effected in a common delay unit, and each output of the common delay unit relates to one propagation path. These outputs are typically referred to as fingers.
In the delayed signal, chips may have been omitted or duplicated to compensate for time drift. The omission or duplication of chips is also controlled by the control pulse. This delayed signal is thereafter supplied to demodulation units in the receiver. These demodulation units are employed to reconstruct the symbol that was initially translated into a number of chips at the transmitter. To these demodulation units, a compensation signal is provided that indicates whether chips have been omitted or duplicated in the delayed signal. This is necessary for the demodulation units to consider the omission or duplication of chips in the delayed signal when demodulating the delayed signal, and thereby make a correct demodulation of the chips.
The present invention is advantageous for a number of reasons, firstly since the variable delays are applied to the chips rather than to the symbols. This greatly decreases the performance loss of the receiver, since a chip is omitted instead of a complete symbol. Only 1/N of a symbol consisting of N chips is omitted/repeated by this operation. The delay adjustments will occur N times more than in the case where the delays are applied on a symbol level, but these N compensations are spread over many symbols, so the performance loss is decreased. Secondly, at the output of the common delay unit, the fingers of the receiver are operated to accomplish time drift compensation, so that the chips that originate from the respective propagation paths are output, i.e. the position of a finger is altered, at the same instant of time. This greatly simplifies the software management of the receiver.
According to an embodiment of the invention, the compensation signal is synchronized with the control pulse. Thus, the frequency of the compensation signal equals the frequency of the control pulse. This is to ensure that the compensation signal occurs as often as the control pulse does. Preferably, on each occasion that time drift compensation is achieved by means of omission or duplication of chips in the received signal, the compensation signal shall notify the demodulation units thereof.
According to another embodiment of the invention, the control pulse is aligned with a symbol boundary. In UMTS, one slot is equal to a sequence of 2560 chips. When no time drift compensation is to be effected in the receiver, the control pulse is activated after the reception of one complete slot, i.e. 2560 chips. However, when time drift compensation is to be undertaken, the control pulse will be activated after 2559 chips of one slot has been received in case of chip omission, or after 2561 chips has been received in case of chip duplication.
According to a further embodiment of the invention, the received signal comprises both data chips and pilot chips and the demodulation is performed in a first and a second demodulation unit. The received pilot chips are demodulated in the first demodulation unit to produce demodulated pilot chips, and the received data chips are demodulated in the second demodulation unit to produce demodulated data chips. The use of a pilot channel enables channel estimation by means of transmitting a predefined chip sequence over the channel. By employing a pilot channel, the receiver can compute an estimate of the channel and take this estimate into account when reconstructing the symbols.
According to yet further embodiments of the invention, the compensation signal is given a first value which indicates that a chip has been omitted in the delayed signal and a second value which indicates that a chip has been duplicated in the delayed signal.
The demodulation of chips is performed by first descrambling the delayed chips and than despreading the descrambled chips. The reconstruction of a symbol is effected by integrating the demodulated pilot chips to create a pilot symbol and the demodulated data chips to create a data symbol. The first and second values of the compensation signal control scrambling codes and spreading codes provided to the demodulation units such that a chip is omitted in the respective code on reception of the first value and a chip is duplicated in the respective code on reception of the second value. The first and second values of the compensation signal further control the integrators such that a chip is omitted in the integration on reception of the first value, and a chip is duplicated in the integration on reception of the second value.
According to still another embodiment of the invention, the first demodulation unit delivers channel estimation information derived from the pilot symbol to the second demodulation unit. The second demodulation unit employs the channel estimation information to improve the data symbol by taking into account channel parameters. The second demodulation unit advantageously performs de-rotation processing on the data symbols based on the channel estimation information from the first demodulation unit. Thereafter, the data symbols are processed by decoders and delivered to higher layer processors. By performing the derotation, the phase rotation caused by the channel can be eliminated.
Further features of, and advantages with, the present invention will become apparent when studying the appended claims and the following description. Those skilled in the art realize that different features of the present invention can be combined to create embodiments other than those described in the following.
The preferred embodiments of the present invention will be described in detail in the following with reference made to the accompanying drawings, in which:
In UMTS, spreading is used at the transmitting side, i.e. at the base station, to translate each symbol into a sequence of bits. Each symbol is multiplied with a spreading code comprising N bits, or chips, per symbol. The number N of chips in the spreading code is typically known as the spreading factor (SF). The value of SF varies from 4 to 256 depending on the data channel, but in case of the pilot channel employed in the mobile phone receiver, SF is set to 256.
Thus, a two-stage modulation technique is employed at the receiver. A prior art receiver is schematically shown in
After having passed the common delay unit 201, the signals are supplied to a first demodulation unit, i.e. a master rake 202 and a second demodulation unit, i.e. a slave rake 203. The receiver 200 shown in
The master rake 202 comprised in each finger demodulates received predefined chips that are transmitted over a pilot channel. First, the pilot chips are descrambled, i.e. demodulated by a scrambling code provided by a master scrambling code generator 204. Second, the descrambled chips are despread, i.e. demodulated by a spreading code provided by a master spreading code generator 205. Thereafter, a first master integrator 206 performs an integration over 256 chips to reconstruct one pilot symbol. A second master integrator 207 performs an integration over four symbols and delivers channel estimation information to the slave rake 203. The slave rake 203 first descrambles chips corresponding to the actual data channel, i.e. the channel associated with the propagation path for which timing compensation is to be undertaken. The descrambling is performed by means of a slave scrambling code generator 208. A slave spreading code generator 209 is then employed to perform the despreading of data chips. A slave integrator 210 performs an integration over SF chips to reconstruct one pilot symbol.
The slave rake performs de-rotation processing on the data symbols based on the channel estimation information from the master rake 202. Thereafter, the data symbols are processed by decoders 211 and delivered to higher layer processors.
The despreading and descrambling code generators 204, 205, 208, 209, the integrators 206, 210 and the delay unit 201 must be synchronized with the received chip sequence in order to correctly reconstruct the symbols of each path. In accordance with the present invention, a control pulse—a slot pulse—is supplied to the delay unit 201, and a first and a second compensation signal is supplied to the master rake 202 and the slave rake 203, and thus to the despreading and descrambling code generators 204, 205, 208, 209 and the integrators 206, 210 comprised in the rakes.
This synchronization is performed by the timer 212, and will be described in detail in the following. Note that it is possible to perform the demodulation without a pilot tone and, thus, without the master rake 202. However, it would in that case not be possible to provide the slave rake 203 with channel estimation information. Consequently, the performance of the receiver 201 would be severely degraded.
In UMTS, following definitions are used:
To perform time drift compensation as provided by the present invention, a symbol number counter must be employed at the receiver to count the number of each symbol in a slot, i.e. 1-10 in the master rake 202 and 1-2560/SF in the slave rake 203. This counter is supplied with the number of the symbol for which compensation is to be achieved.
Note that the symbol for which compensation is accomplished in a slot is not necessarily symbol number 10 (for the master rake 202) and symbol number 2560/SF (for the slave rake 203). In UMTS, several channels are received simultaneously. These channels can be delayed in relation to each other, and the symbol boundaries of these delayed channels will be time-aligned, but a symbol can be shifted relative to other channels. This is illustrated in
As mentioned hereinabove, when a slot extension is effected, there is an “extra” chip in the corresponding slot. This extra chip is a result of the shift of fingers in the common delay unit 201 and shall not be considered. This is done by limiting the integration 206, 210 in the demodulation operation to operate on 256 chips (for the master rake 202) and 2560/SF chips (for the slave rake 203). However, the integrators 206, 210 shall wait for the duration of one chip, which duration corresponds to the extra chip, before demodulation of the next symbol is performed. When a slot extension is effected, the last chip will be repeated and thus the integrators 206, 210 will sum that last chip twice.
When a compensation is to be performed in accordance with the present invention, i.e. a chip omission or a chip repetition is effected, a compensation signal which is synchronized with the slot pulse is supplied to the code generators 204, 205, 208, 209 by the timer 212. The compensation signal has a first value to indicate a chip omission, and a second value to indicate a chip repetition. The compensation signal is also supplied to the integrators 206, 210 to ensure a correct integration operation. The timer 212 synchronizes the compensation signal with the slot pulse such that they are submitted to generators and integrators concerned at the same instant of time that the slot pulse is submitted to the delay unit 201.
Thus, to perform time drift compensation in accordance with the present invention, a number of resources are required in the receiver of
For slot reduction, a slot pulse must be generated which is located at a distance of 2559 chips from the last slot pulse. Slot pulses are generated by the timer 212 and sent to the common delay unit 201, such that the delay of each finger is synchronized with the slot pulse as shown in
For slot extension, a slot pulse must be generated which is located at a distance of 2561 chips from the last slot pulse. The timer 212 generates slot pulses and sends them to the common delay unit 201, such that the delay of each finger is synchronized with the slot pulse as shown in
Even though the invention has been described with reference to specific exemplifying embodiments thereof, many different alterations, modifications and the like will become apparent to those skilled in the art. The embodiments described are therefore not intended to limit the scope of the invention, as defined by the appended claims.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4977578 *||Feb 21, 1989||Dec 11, 1990||Victor Company Of Japan, Ltd.||Spread spectrum communication system|
|US5684828 *||Oct 10, 1995||Nov 4, 1997||Dallas Semiconductor Corp.||Wireless data module with two separate transmitter control outputs|
|US5856998 *||Dec 18, 1996||Jan 5, 1999||Omnipoint Corporation||Method and apparatus for correlating a continuous phase modulated spread spectrum signal|
|US6282234||Apr 18, 2000||Aug 28, 2001||Sharp Kabushiki Kaisha||Spread spectrum receiver|
|US6377606||Mar 24, 1997||Apr 23, 2002||Nokia Telecommunications Oy||Method for reducing interference, and a radio system|
|US20040184513 *||Jul 23, 2003||Sep 23, 2004||Stein Lundby||Method and apparatus for multipath demodulation in a code division multiple access communication system|
|U.S. Classification||370/350, 375/149, 375/148|
|International Classification||H04B1/707, H04B1/00, H04J3/06|
|Cooperative Classification||H04B1/7085, H04B1/7117, H04B1/7115|
|Jun 22, 2006||AS||Assignment|
Owner name: KONINKLIJKE PHILIPS ELECTRONICS N.V.,NETHERLANDS
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:ARDICHVILI, EMMANUEL;GANDER, MARTIAL;REEL/FRAME:018072/0051
Effective date: 20060330
|Aug 17, 2007||AS||Assignment|
Owner name: NXP B.V.,NETHERLANDS
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:KONINKLIJKE PHILIPS ELECTRONICS N.V.;REEL/FRAME:019719/0843
Effective date: 20070704
|Oct 24, 2013||FPAY||Fee payment|
Year of fee payment: 4